By Josh Perry, Editor
Researchers at the U.S. Department of Energy (DOE) Argonne National Laboratory (Lemont, Ill.) developed a method for creating the molecular structures of materials for next-generation organic electronics by using machine learning.
Schematic of the ANN-ECG method used in this work. (Argonne National Laboratory)
According to a report from the lab, this new method drastically speeds up the process of screening potential organic materials for electronics. Current manufacturing processes require intricate details and a shift in conditions could drastically alter the final material’s properties.
In this method, vapor deposition is the main assembly technique, which slowly produces thin films of a material on the surface of a substrate with scientists capable of fine-tuning the layering of the molecules to impact the material’s charge mobility.
“To simulate the packing of entire devices, often containing millions of molecules, scientists must develop computationally cheaper, coarser models that describe the behavior of electrons in groups of molecules rather than individually,” the article explained. “These coarse models can reduce computation time from hours to minutes, but the challenge is in making the coarse models truly predictive of the physical results.”
This is where machine learning comes into play. The simulations predict the electronic properties of the material using coarse models. This allows researchers to examine significantly more molecular arrangements than before.
The research was recently published in Science Advances. The abstract read:
“Computational studies aimed at understanding conformationally dependent electronic structure in soft materials require a combination of classical and quantum-mechanical simulations, for which the sampling of conformational space can be particularly demanding.
“Coarse-grained (CG) models provide a means of accessing relevant time scales, but CG configurations must be back-mapped into atomistic representations to perform quantum-chemical calculations, which is computationally intensive and inconsistent with the spatial resolution of the CG models.
“A machine learning approach, denoted as artificial neural network electronic coarse graining (ANN-ECG), is presented here in which the conformationally dependent electronic structure of a molecule is mapped directly to CG pseudo-atom configurations.
“By averaging over decimated degrees of freedom, ANN-ECG accelerates simulations by eliminating backmapping and repeated quantum-chemical calculations. The approach is accurate, consistent with the CG spatial resolution, and can be used to identify computationally optimal CG resolutions.”